Reason: geodetic more closely matches record.
However, one must have to assume that the record reference direction is True North. How did the other guy determine this? Solar observations by the hour angle method are around +- 10 seconds or so. Older surveys were maybe a Polaris observation, or a solar observation using the altitude method. Or maybe using a Roloff prism in a theodolite or aiming a one minute transit at the sun and looking at the cross hairs on a sheet of paper held behind the instrument. Or even using a magnetic compass declinated to true north, which is probably the most common method utilized for the USPLSS around here. How close can one read a magnetic needle on a background graduated in degrees? Maybe to a quarter of a degree, or 15 minutes? For retracement work, it is what it is between found points. For putting back in a section corner or quarter section corner, one must come from found and proven monuments, so the final bearing is still irrelevant. It is what it is. In the extremely rare case of going north 00 degrees, 00 minutes and 00.00 seconds east for 5,280.000 feet to set the section corner, how close is close enough? One second is one foot in forty miles. More important, where did the first surveyor set the section corner? A very high percentage of retracement work means that we look for existing evidence of where the old section lines and corners are. There has to be some shred of evidence along the way to prove that the line is "here" or the old corner is "there." And, if you do enough work in an area, you can develop a scale factor for a particular USPLSS deputy surveyor to better retrace his surveys.
Hey, surveying ain't easy. Nor is it simple. Read, study, try and learn. Make your system talk to you and push it to see when and how it fails you. What really goes on inside that magic box with all them buttons we push?! B-)
I see why a lot of people just use grid directions. How close is close enough? Grid users are just reporting an easily repeatable basis of bearing on their survey plats. It just bothers me if an old deed says "east" and I cannot find the point and must replace it. I have now put back in the ground a point that is not in the same hole that the first survey stake was in.....it is off by five minutes or so in direction around here. Again, how close is close enough? One minute is 0.03 feet per one hundred feet. A 1/2" re-bar is 0.04 feet across.
tds survey pro grid north with ground distanceb
At the tolerances you are working with, Laplace is irrelevant, since in most places it is only a few seconds of arc.
But just to keep things clear, I believe the Laplace correction is only involved in bringing a bearing from some vertical angle to another, and has nothing to do with grid versus ground.
It is most often applied in finding azimuth astronomically. For precise control work in mountainous territory it might also be needed for total station work on the ground.
tds survey pro grid north with ground distanceb
I've scratched my head over exactly what a LaPlace Correction is, and when and how you would ever need to use it.
Dave
tds survey pro grid north with ground distanceb
My interpretation of that figure is:
You are standing on the (brown) ground at the intersection with the red and green straight lines. The geodetic vertical line perpendicular to the ellipsoid model (passing near center of the earth but not exactly) is red. The geoid or level surface is the green wavy line (much exaggerated) and the local gravity vector is the green straight line.
So when you sight something north of you (behind the page) and then raise to a high vertical angle you moving in the plane with the green local vertical (plane also perpendicular to the page). You are departing from the plane of the red line toward the center of the ellipsoid upon which geodetic measurements are based. Laplace lets you calculate how much that affects the azimuth you measure.
That isn't the ideal diagram for illustrating this, but I don't have a better one at hand.
Hope this helps.
tds survey pro grid north with ground distanceb
Bill,
I was following you OK until "So when you sight something north of you..."
In order to use a LaPlace Correction you tilt your gun (Trunnion axis?) to the East or West?
Does a LaPlace Correction tilt your map projection so that your projection is not perpendicular to the Ellipsoid? How would that work?
I'm lost.
Dave
tds survey pro grid north with ground distanceb
Maybe somebody else has a better explanation.
You wouldn't tilt your gun because you don't have calibrated tilt. That's an interesting way to think about it, but not something the textbooks mention doing in practice. The correction is usually applied to the azimuth after taking the data.
Imagine standing at that point and pointing your telescope at something straight behind the screen/page. That direction is geodetic north. The diagram page is oriented west-east.
Then as you point higher up, say to Polaris, your telescope is moving in a plane that is perpendicular to the page and contains the green line. That is where the departure from geodetic north occurs, and is calculated by Laplace.
The Laplace concept doesn't tie directly to projections. You may need Laplace to reduce your observations to geodetic. Going from geodetic to a projection is then a separate subject.
tds survey pro grid north with ground distanceb
Bill,
I've spent a few hours now reading about it, but understanding eludes me:
Sadly, I think that this is one of those things I don't have the brainpower to understand. It's not the first time. I'll just add it the list of things it's not possible for me to understand:
1. The Theory of Relativity
2. Women
3. LaPlace Correction
Dave
Laplace - last try
Try this experiment, in thought or practice, as another way to get the concept of why there is a Laplace correction.
Pick a tall pole and set up your transit/theodolite/total station straight south of it, at a distance where the top of the pole makes a sizable vertical angle from your location, and where you are on the side of a slight hill.
Pretend that somebody set the pole exactly perpendicular to the ellipsoid surface (really, it's probably pretty close) as if it were hidden some distance behind the red line in the figure. The pole marks geodetic north from your location, whether you sight the top or bottom of the pole.
Now, pretend that the local gravity pulls your plumb bob several inches to the side (in reality uphill) from where it really wants to point, and hold it there. This is the green line, and is just like the real gravity vertical might be, if exaggerated to degrees instead of seconds so you can see what it looks like.
You are pretending the ground on your hill is the parallel to the geoid surface (redraw their brown line so the hill continues to go up to the right), and the plumb bob would want its string to be perpendicular to the (pretend) geoid so it has to follow the green line.
Set up the instrument about level in reality, and then crank the tribrach screws so the optical plummet points where we're holding the plumb bob in the pretend gravity. That would make the instrument about level in the pretend gravity. The bubble will be more than maxed out, but ignore it - it doesn't know about the pretend gravity in this experiment.
Find the horizontal angle to the base of the pole. Raise the scope and take the horizontal angle halfway up the pole. Raise it again and take the horizontal angle to the top of the pole. It's a different "horizontal" angle (horizontal in our pretend world) in each case because the pretend gravity and the ellipsoid normal have an angle between them. In our thought experiment it's degrees different, but in reality maybe seconds different.
Any better, or just more mud?
Harold,
I might not understand, or maybe I am doing something wrong, but your procedure sounds very complicated.
I am going to do some studying, and it will probably click. I am not, however, sure why you would need to rotate numerous times until the job is complete and you are ready to draw the plat.
My normal procedure is to go out, collect my data, and rotate my deeds to my raw field data look for monuments, and proceed with the survey. I will collect all of my data for the job, and rotate at the end, when I get ready to draw the plat. If I have to return to set corners, I will upload my rotated info to the data collector, and tie into my control, and set the needed corners.
Most of my jobs are smaller boundaries and topo surveys for engineering design.
Thank you for sharing your procedure. I look forward to learning more about your procedure. I am sure you are on to something.
Have a great weekend.
Jimmy
Laplace - last try
Bill,
"Any better, or just more mud?"
Mud, but I appreciate your efforts. Thank you.
I get the picture, roughly, and I understand the Geoid is lumpy and all. What I don't understand is how the LaPlace Correction works. From what I read, the LaPlace Correction relates physical position to time, and vice-versa. (Whenever somebody starts throwing Greek Letters at me, I'm in trouble.) I don't think LaPlace specifically intended his work to be applied to surveying. Who was it that first thought, "Wow, I can use this to relate Astronomic observations to Geodetic!" What I'd really like to know is if GPS measurements make LaPlace Corrections irrelevant, or do those corrections have a place somewhere in figuring a bearing.
Dave
Laplace - last try
> Bill,
>
> "Any better, or just more mud?"
>
> Mud, but I appreciate your efforts. Thank you.
>
> I get the picture, roughly, and I understand the Geoid is lumpy and all. What I don't understand is how the LaPlace Correction works. From what I read, the LaPlace Correction relates physical position to time, and vice-versa. (Whenever somebody starts throwing Greek Letters at me, I'm in trouble.) I don't think LaPlace specifically intended his work to be applied to surveying. Who was it that first thought, "Wow, I can use this to relate Astronomic observations to Geodetic!"
Perhaps the view from the unique perspective of a grasshopper will help...
Here goes:
Sculpt the earth precisely as it is in the real world out of wax. Every mountain, every valley. Now, spin it on a stick and put a blow torch on it to start melting the wax...into a lumpy, but smoother surface. That's the GEOID. Your plumb bob will be perpendicular to that no matter where you are on the surface of the actual earth.
But there are so many lumps and bumps in that surface that there's no computer on earth that can model it mathematically. Satellite GPS systems need an accurate, MATHEMATICAL surface to base their measurements on.
So the brains above our pay grade CREATED one...not perfect, but close, mathematical surface...That's the "reference Ellipsoid".
The LaPlace correction calculates the difference at any given point between the Geoid and the Ellipsoid (Time has nothing to do with it as you infer above, just location). I could be wrong on that last point. I'm not sure I've ever seen the application of such things as the lunar orbit effect on gravity...but I don't think it's significant to our measurements.
Anyway, LaPlace calculates the difference between which direction your plumb bob is telling you "UP" is, and where GPS is telling you it is.
>What I'd really like to know is if GPS measurements make LaPlace Corrections irrelevant, or do those corrections have a place somewhere in figuring a bearing.
>
> Dave
If you are relying 100% on GPS, then I guess LaPlace does not come into play, because it uses the reference Ellipsoid.
But if you mix measurements...Anything that uses a plumb bob or an optical plummet (which is gravity based), and GPS, well, then, you need to use the LaPlace correction.
Laplace - last try
RFC,
Thank you. I'm going to quit hijacking Mark's thread and start my own.
Dave
Laplace - last try
No, there is nothing about time in the Laplace calculation. Time is used in calculating position or azimuth from astronomical bodies. Once you have an astronomical azimuth, you are done considering time. Converting to geodetic is where Laplace comes in, without regard to time.
There isn't any need for a Laplace correction when making GPS measurements.
However, there is still a possibility that you might need Laplace correction for doing precise total station work, to correct measured horizontal angles between stations at very different vertical angles. (That's why you need it in astro work.)
Laplace - last try
Bill,
I was just going by this Wikipedia Link:
"The Laplace transform is a widely used integral transform in mathematics and electrical engineering named after Pierre-Simon Laplace (/l??pl??s/) that transforms a function of time into a function of complex frequency."
Dave
Laplace - last try
Mr. Laplace did pioneering work in a lot of mathematical areas. The fact that his name is on it doesn't mean it's closely tied to something else with his name on it.
The geophysical Laplace Correction is totally unrelated to the Laplace Transform.
Josh,
That's amazing! We did the same practice when we used TDSvSurveyPro. Carry grid cords in DC, but staked ground distances for layout. I thought it was pretty clever, until we moved to the Leica DC with Carlson SurvCE. I called Carlson one day and asked them where that setting was, to toggle grid-to-ground stakeout using the scale factor?
After the 30 minute lecture/shellacking I got from the good folks at Carlson as to how much a baffoon I must be, and how that kind of practice is unprofessional, etc, etc., I just let it drop. We had bigger fish to fry, as we were still learning the new robot gun and collector.
The guy on phone support made a couple good points, but I don't know. One of the points I couldn't argue with: We work on the grid (approximate, since we're an autoCAD shop,) and all our designs were grid distances. Say for example we need to stake a 600 ft by 300 ft building pad. Those 600x300 dimensions were grid in the autoCAD model. Then I stake them with a 1.000000345 scale factor, and the actual pins I set in the footings will be 600+ feet apart on the ground. Check. Now, say the concrete guy comes up behind me with his EDM, and measures between my mag nails that I set in the footing. He comes up, not with 600.00, but rather the long distance, at that point the whole job site starts puckering up, and they all eventually come to the conclusion that I don't know what I'm doing.
Again, the jury is still out on this topic.
Yes, there is a difference as you knew. The greater the vertical distance is between grid and ground, the greater the horizontal difference at ground level will be. Erecting concrete structures and putting together prefabricated steel structures on grid layout is not good - things don't match exactly when check measurements are made using conventional construction measurement tools. A contractor who checks your work will have slightly different measurements and question your ability. And yes, It is not good for the reputation. It would seem that using a ground-based coordinate system would work best for engineering projects with critical points. (After a re-read of your post, I realized that your design work and construction layout was on grid. But actual construction was on ground, and the vertical separation was large enough to cause a difference.)
An old-school way to look at it would be to buy a brand new 100-foot steel chain with a defect that made the chain, say, 100.01 feet long under ideal conditions at the 100-foot mark. You, of course, did everything right and applied the proper chaining tension and adjusted for the temperature correction factor. Your layout points are indicated at 100.00 feet apart according to your chain, but the guy behind you checks your work with his chain and gets 100.01 feet. Obviously, to him, you are at fault.
Knowing there is a difference between grid and ground and then applying the proper procedure to utilize a derived correction factor for ground layout using GPS is something those in the construction surveying industry will need to do in order to make GPS points match ground based construction measurements. Even in land surveying, a section line around here that is one mile long will have a ground measurement of, say, 5,280.46 feet if it was laid out at 5,280.00 grid.
Survey Pro has a job set-up checkbox where a projection system will not be used. One must occupy existing points to calibrate the GPS ground-based coordinate system to the existing coordinate system. I have tried that on a few old surveys taking shots on old points with relative success. I get some slight differences since all my original elevations are "100" or zero. I use State Plane Coordinates for my land surveying projects. Since I have only one coordinate system, all my jobs are, in fact, "georeferenced." I can dump my section corners, control points, corners found and set and my boundaries into some GIS software and be able to overlay that information on aerial photos or the old USGS quad maps. I can do mission planning and utilize the database to store job information.
I have not done much engineering surveying with my GPS system. I believe that I would adopt a no-projection or a low distortion projection system if I do. I would be interested to know how the construction surveying guys utilize GPS to do their work.
The guy on phone support made a couple good points, but I don't know. One of the points I couldn't argue with: We work on the grid (approximate, since we're an autoCAD shop,) and all our designs were grid distances. Say for example we need to stake a 600 ft by 300 ft building pad. Those 600x300 dimensions were grid in the autoCAD model. Then I stake them with a 1.000000345 scale factor
Heck at that scale factor you are basically surveying right on the ground with grid anyway. I wouldn't do anything if that is the factor. 0.01' won't show up until you lay out a line 30,000' long.
